(PLA2R1) Sequence Variants in Idiopathic Membranous Nephropathy

CLINICAL RESEARCH www.jasn.org

Phospholipase A2 Receptor (PLA2R1) Sequence Variants in Idiopathic Membranous Nephropathy

† ‡ Marieke J.H. Coenen,* Julia M. Hofstra, Hanna Debiec, Horia C. Stanescu,§ | Alan J. Medlar,§ Bénédicte Stengel, ¶ Anne Boland-Augé,** Johanne M. Groothuismink,* †† ‡‡ Detlef Bockenhauer,§ Steve H. Powis,§ Peter W. Mathieson, Paul E. Brenchley, † ‡ Robert Kleta,§ Jack F.M. Wetzels, and Pierre Ronco

*Department of Human Genetics and †Department of Nephrology, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands; ‡Institut National de la Santé et de la Recherche Médicale, Université Pierre et Marie Curie Univ-Paris, Assistance Publique-Hôpitaux de Paris, Department of Nephrology, Tenon Hospital, Paris, France; §Centre for Nephrology, University College London, London, United Kingdom; |Institut National de la Santé et de la Recherche Médicale U1018, CESP, Team 10, Villejuif, France; ¶Université Paris-Sud, Villejuif, France; **Centre National de Génotypage, Institut de Génomique CEA, Evry, France; ††Academic Renal Unit, University of Bristol, Bristol, United Kingdom; and ‡‡School of Biomedicine, University of Manchester, Manchester, United Kingdom

ABSTRACT The M-type receptor for phospholipase A2 (PLA2R1) is the major target antigen in idiopathic membranous nephropathy (iMN). Our recent genome-wide association study showed that genetic variants in an HLA- DQA1 and phospholipase A2 receptor (PLA2R1) allele associate most significantly with biopsy-proven iMN, suggesting that rare genetic variants within the coding region of the PLA2R1 gene may contribute to antibody formation. Here, we sequenced PLA2R1 in a cohort of 95 white patients with biopsy-proven iMN and assessed all 30 exons of PLA2R1, including canonical (GT-AG) splice sites, by Sanger sequencing. Sixty patients had anti-PLA2R1 in serum or detectable PLA2R1 antigen in kidney tissue. We identified 18 sequence variants, comprising 2 not previously described, 7 reported as rare variants (,1%) in the Single Nucleotide Polymorphism Database or the 1000 Genomes project, and 9 known to be common polymorphisms. Although we confirmed significant associations among 6 of the identified common variants and iMN, only 9 patients had the private or rare variants, and only 4 of these patients were among the 60 who were PLA2R positive. In conclusion, rare variants in the coding sequence of PLA2R1, including splice sites, are unlikely to explain the pathogenesis of iMN.

J Am Soc Nephrol 24: 677–683, 2013. doi: 10.1681/ASN.2012070730

Idiopathic membranous nephropathy (iMN) is the established that 70% of white patients with active leading cause of nephrotic syndrome in the adult disease have antibodies against PLA2R1.1,3,4 white population. It is characterized by immune complex deposition in the subepithelial layer of the glomerular basement membrane. On the basis of Received July 25, 2012. Accepted December 5, 2012. studies in Heymann nephritis, a model of MN in the M.J.H.C., J.M.H., and H.D. contributed equally to this work; R.K., rat, it was suggested that immune complexes were J.F.M.W., and P.R. contributed equally to this work formed locally, by the binding of antibodies to a Published online ahead of print. Publication date available at 1 podocytic antigen. The description of neonatal www.jasn.org. MN caused by maternal antibodies against neutral Correspondence: Dr. Marieke Coenen, Department of Human 2 endopeptidase provided proof of concept. Further Genetics, Radboud University Nijmegen Medical Centre, PO Box evidence was provided by the discovery of the 9101, 6500 HB Nijmegen, The Netherlands. Email: m.coenen@ M-type receptor for phospholipase A2 (PLA2R1) as gen.umcn.nl the predominant autoantigen in iMN.3 It is well Copyright © 2013 by the American Society of Nephrology

J Am Soc Nephrol 24: 677–683, 2013 ISSN : 1046-6673/2404-677 677 CLINICAL RESEARCH www.jasn.org

Although no studies have firmly proven that the antibodies are these rare variants (rs141800672, rs149256089, rs150221555, indeed pathogenic, genetics studies have supported the impor- rs140427239, rs149960520, rs149133741, and rs181959329) tant role of PLA2R1 in the pathogenesis of iMN. Recently, we have been reported in publicly available data sets with very have shown highly significant associations to iMN with single low minor allele frequencies (#0.2%). We identified two nucleotide polymorphisms (SNPs) located on chromosomes 6 novel missense variants that have not been previously reported, and 2, by means of genome-wide association studies (GWAS) which result in amino acid changes (c.1160G.A, p.Arg387His; in three independent cohorts.5 The region of interest defined c.2060T.C, p.Leu687Pro). The novel variant c.1160G.A, on chromosome 2 includes the gene PLA2R1. SNPs in this gene p.Arg387His and the splice site change c.3850+1G.A (IVS26 have also been shown to be associated to iMN susceptibility in +1G.A) (rs181959329) were carried by the same patient. two candidate gene association studies performed in Korea and Additionally, nine common SNPs were observed; six of them Taiwan.6,7 These findings together suggest an important role for in the coding regions (rs4665143, rs3749117, rs35771982, PLA2R1 in the pathogenicity of iMN. rs33985939, rs72954858, and rs3828323) and three in noncod- Our GWAS was based on common SNPs, as is typical for ing regions (rs2715918, rs925409, and rs3749119). most GWAS. These SNPs serve as genetic markers used to In the subgroup of 60 anti-PLA2R-positive patients, four identify alleles or haplotypes of interest. Of note, our GWAS of the rare variants (rs149256089, rs140427239, rs149960520, study showed that both the HLA locus on chromosome 6 and and rs149133741) were each observed once in four different the PLA2R1 locus on chromosome 2 were significantly asso- patients. The two novel variants were not observed in this ciated with iMN not only on the SNP level (basic allele test) subgroup (Table 2). but also on the haplotype level (haplotype association test).5 Of the 15 observed coding variants, 12 (rs149133741, These SNPs demarcate alleles that are different in cases versus rs149960520, rs140427239, rs33985939, rs150221555, controls. Not only common SNPs but also rare mutations can rs35771982, rs3749117, rs149256089, rs141800672, rs4665143, be the cause of a signal identified by GWAS.8 Although iMN is and the novel p.Arg387His and p.Leu687Pro) are located with- not a simple genetic disorder with classic Mendelian inheri- in the regions coding for known domains of PLA2R1, of which 1 tance, the strong and significant association of this phenotype (rs3749117) is located within the WMGL motif of the C-type to PLA2R1 (and the HLA locus) leads to the hypothesis that lectin-like domain (CTLD)-1 of PLA2R1. Two other coding rare genetic variants within the coding region of the PLA2R1 variants, one common (rs33985939, p.Arg404His) and one rare gene may explain antibody formation. In the current study, we (rs140427239, p.Tyr499Cys), are situated one and two amino performed sequencing of the coding regions of the PLA2R1 acids, respectively, away from the cysteine residues involved in gene to confirm this hypothesis. disulfide bond formation (C4 and C1 in CTLD2; see Table 1 and Figure 1). The essential splice site at the border between exon 26 and RESULTS intron 26 is altered in one patient, c.3850+1G.A(IVS26 +1G.A) (rs181959329), predicted to lead to exon 26 being Patient Characteristics skipped, which is in frame. This leads to a shortened PLA2R1 Ninety-five patients with biopsy-proven iMN were included: protein missing 46 amino acids within CTLD8, obliterating 48 from the French cohort and 47 from the Dutch cohort, the first cysteine residue involved in disulfide bond formation respectively. Patients were mostly male (75%), and the mean (C1). Of note, this is a known sequence variant with a yet un- age 6 SD at diagnosis was 51615 years. All patients were of determined allele frequency in the general population. self-reported white ancestry. Of note, Single Nucleotide Polymorphism Database release Serum samples were available for 82 patients; anti-PLA2R 136 (February 2012) currently has 138 human sequence variants antibodies were present in 43 samples. In addition, 17 patients curated within the coding region, including splice sites of were considered positive according to the presence of PLA2R PLA2R1. Actually, only 8 of the 138 variants are frequently ob- antigenintheirkidneybiopsyspecimen(SupplementalTableS1). served, nonsynonymous, coding variants. Minor allele frequencies were not established for most of those SNPs. Sequencing All 30 exons, including essential splice sites (defined as the two Association intronic nucleotides—canonical GT and AG—at the intron- The common SNPs identified by our sequencing permit us to exon boundaries, respectively) were sequenced by Sanger tech- analyze the PLA2R1 region in more detail compared with the nology. We identified 18 variants, including 3 in noncoding GWAS that mainly included common intronic SNPs (typical regions (Table 1). for SNP chips used for GWAS). Of the 18 variants observed, 9 were rare or novel (see the In the 95 patients with iMN studied, the allele frequency of Concise Methods section for definitions). These 9 variants several SNPs was significantly different from that of the ethni- were all encountered in a heterozygous state in one patient cally matched controls from the 1000 Genomes project (see each, except for the rare variant rs149133741 (c.2038G.T, Table 1). Specifically, there was a significant difference for the p.Val680Leu), which was observed in two patients. Seven of coding SNPs rs4665143 (p.Ser87Ser), rs3749117 (p.Met292Val),

678 Journal of the American Society of Nephrology J Am Soc Nephrol 24: 677–683, 2013 www.jasn.org CLINICAL RESEARCH

rs35771982 (p.His300Asp), and rs3828323 (p.Gly1106Ser). In addition, two of the three noncoding er; ref, ﬁ SNPs (rs2715918, rs3749119) were signiﬁcantly associated with iMN. The nonsynonymous SNP

value, comparison of (rs3749117; p.Met292Val) located in P the WMGL motif of CTLD1, which has been shown to be in linkage disequilibrium (r2=0.70) with the most significantly associated SNP located in the first intron from our previous GWAS (i.e., rs4664308; 2 P=4.3310 28) with 556 iMN cases, 26 Donor ss was significantly associated with 2 iMN in this study (P=3.1310 6). SNP rs4665143 was significantly as- 2 sociated with iMN (P=2.1310 5), -UTR 1 -UTR 1 9 9 fi 5 Ess con rming our previous work. Two nonsynonymous SNPs identified in this study (rs35771982 [p.His300Asp] and rs3828323 [p.Gly1106Ser]) were also assessed by the two candidate gene association A C Intron 16 6 .

. studies from Taiwan(p.His300Asp) and T p.Val680Leu Missense 13 CTLD4 A p.Gly1106Ser Missense 24 Linker region between CTLD6 and CTLD7 A p.Pro600= SynonymousA 11 p.Arg387His CTLD3 Missense 7 CTLD2 A p.Arg404His Missense 7 CTLD2 1 aa from C1 T p.His1105= Synonymous 24 Linker region between CTLD6 and CTLD7 T p.Arg387Cys Missense 7 CTLD2 G p.Tyr499Cys Missense 9 CTLD2 2 aa from C4 C p.Leu687Pro Missense 13 CTLD4 A5 T p.Asp200Tyr Missense 2 FNII G p.His300Asp Missense 5 CTLD1 G p.Met292Val Missense 5 WMGL motif of CTLD1 C p.Ser87= Synonymous 2 CRD C p.Asp272= Synonymous 4 CTLD1 . . . . . T5 . . . 7 . .

. Korea (p.His300Asp, p.Gly1106Ser) . . . . . and found to be signiﬁcantly associated with iMN. Both SNPs is on the negative strand); cDNA, NM_007366.4; prot level, NP_031392.3; Ess, essential splice site; ss, splice reached signiﬁcance for allelic association with iMN in the present study T c.2060T T c.816T A c.1496A C c.1159C C c.3315C G c.1800G G c.1160G G c.598G G c.3850+1G G c.2038G G c.1211G 23 PLA2R1 (P=1.6310 for rs3828323 and 2 P=5.7310 6 for rs35771982). These associations were also clearly present when we limited the analysis to the Value Ref cDNA cDNA Prot Level Effect Exon Observations

P anti-PLA2R1–positive subgroup (Table 2).

0.005 0.005 0.005 0.006 0.005 0.005 DISCUSSION d In a series of three independent GWAS observed in the present study (95 patients with iMN) d d d d d in biopsy-proven cases of iMN with similar genetic background (all of

bronectin type II domain; CRD, cysteine-rich domain; UTR, untranslated region. white ethnicity), as well as in a com- PLA2R1 ﬁ T T/C 0.485 0.296 0.0000031 A c.874A A A/G 0.821 0.716 0.0013 T c.2437+8T A A/G 0.629 0.454 0.000021 T c.261T C C/T 0.511 0.382 0.0016 G c.3316G C C/T 0.339 0.186 0.000049 G c.-106G G G/C 0.480 0.296 0.0000057 C c.898C bined analysis of all three populations, we were able to show that HLA-DQA1 b b b b b b and PLA2R1 haplotypes are associated with iMN with high levels of statistical signiﬁcance, with a combined odds ratio of approximately 80 for individuals homozygous for both risk alleles.5 Sequence variants in We and others have speculated that

cantly associated with iMN. rare sequence variants in PLA2R1, ﬁ 160803929 rs181959329 C C/T 0160840562 Novel160840584 rs149133741160862197 C 0.005 rs149960520160873180 A/C C rs140427239 A 0.005 C/T T A/G160879310 0 0 C/T Novel160879311 0.011 0 rs150221555 G A/G C 0 160889495 C/T rs149256089 0.005 160898605 0 A rs141800672 A/G C 0.001 A/C 0 not studied by GWAS, could cause c c a a a a a a a Signi Not reported in 1000 Genomes; allele frequency from Single Nucleotide Polymorphism Database, release 136. 2 2 2 160885442 rs3749117 2 160808075 rs3828323 2 160918984 rs925409 T T/A 0.074 0.080 0.78 A c.-70A 2 160879259 rs33985939 C C/T 0.094 0.041 0.02 22 160808076 rs72954858 160833188 rs2715918 G G/A 0.012 0.027 0.13 2 160885418 rs35771982 2 160919020 rs3749119 2 160901517 rs4665143 2 2 2 2 2 2 2 Rare variants. Novel variants. 5,9 Table 1. Chr Position Name Ref All CEU AF AF Patients Chr, chromosome; position, base pair position based on University of California, Santa Cruz, genome browser versiona human (February 2009) (GRCh37/hg19)b assembly NC_000002.11; name, rsc identi d site; aa, amino acid; WMGL, motif; FNII, genomic reference allele; all, alleles;variation CEU allele AF, frequencies variation in controls allele and frequencies patients; ref white cDNA, population reference allele from cDNA 1000 (given Genomes that release 10 (March 2012); AF patients, patient variation allele frequencies; the disease. Under this hypothesis,

J Am Soc Nephrol 24: 677–683, 2013 PLA2R1 Sequence Variants in iMN 679 CLINICAL RESEARCH www.jasn.org

the causal or predisposing variants of the PLA2R1 sequence would modify the structure of the protein to give rise to a pep- ﬁ

er; ref, tide/epitope that would t the antigen-presenting groove of a ﬁ complementary, also causal/predisposing variant to HLA genes (e.g., HLA-DQA1). Such PLA2R1 variants might not have been covered in the earlier GWAS because of the design of the geno-

value, comparison of typing chip (which does not include all possible SNPs) or be- P cause the variant was very rare in the general population but not rare in the cohort with biopsy-proven iMN. Our study was designed to assess these possibilities and to shed further light on the genetics of iMN. We sequenced the PLA2R1 gene in 95 patients with iMN, including 60 patients with PLA2R-related iMN as assessed by anti-PLA2R1 antibodies and PLA2R1 antigen in immune deposits. We made two important observations. First, we found

hg19) assembly NC_000002.11; name, rs identi nine rare variants, but in only nine patients, all in a heterozy-

s, patient variation allele frequencies; gous state, and only two of these had not been reported previously (potential “private mutations”). When we focused on -UTR 1 -UTR 1 9 9 the 60 patients with established PLA2R1-related MN, no “private mutations” were found, and four of the rare variants were present. It is thus evident that iMN is not caused by classic Mendelian mutations in the coding region, including splice sites of PLA2R1. Moreover, these results do not support the hypothesis that rare variants present in the coding region of C Intron 16 . T p.Val680Leu Missense 13 CTLD4 A p.Gly1106Ser Missense 24 Linker region between CTLD6 and CTLD7 A p.Pro600= Synonymous 11 CTLD3 A p.Arg404His Missense 7 CTLD2 1 aa from C1 T p.His1105= Synonymous 24 Linker region between CTLD6 and CTLD7

G p.Tyr499Cys Missense 9 CTLD2 2 aa from C4 PLA2R1 are responsible for the association with iMN. These A5 G p.His300Asp Missense 5 CTLD1 G p.Met292Val Missense 5 WMGL motif of CTLD1 C p.Asp272= Synonymous 4 CTLD1 C p.Ser87= Synonymous 2 CRD . . . . T5 . . . ﬁ . . . . ndings agree with the rarity of familial forms of iMN, which . could still be related to “private mutations.” Individual rare ﬂ

is on the negative strand); cDNA, NM_007366.4; prot level, NP_031392.3; aa, amino acid; WMGL, motif; mutations in the coding region of PLA2R1 might in uence antigen processing in only a minority of individuals with iMN, but data providing additional insights are lacking. Al- PLA2R1 T c.816T A c.1496A A c.-70A G c.1800G G c.2038G though our data allow us to make a clear-cut statement that positive patients

– neither rare variants nor mutations in the coding region of the PLA2R1 gene explain the proven genetic association within the PLA2R1 locus, we identiﬁed a few single rare variants in Value Ref cDNA cDNA Prot Level Effect Exon Observations

P the coding region of PLA2R1 in our population. The second important observation is that we identified six common variants in the coding regions of PLA2R1 and three more in the exon/intron boundaries (Table 1). We found a 0.008 0.008 significant association between iMN and four coding and two noncoding SNPs, thus confirming that the PLA2R1 region c is associated with iMN.5 Of note, three significantly associated observed in 60 anti-PLA2R1 c coding SNPs were nonsynonymous. One is the nonsynony- 2 mous SNP (rs3749117; p.Met292Val; P=3.1310 6) located in the WMGL motif of CTLD1, which was previously shown to PLA2R1 2 T T/C 0.485 0.250 0.0000021 A c.874A A A/G 0.821 0.700 0.0020A T A/G 0.629 c.2437+8T 0.430 0.000049 T c.261T C C/T 0.511 0.381 0.0090 G c.3316G C C/T 0.339 0.152 0.000049 G c.-106G G G/C 0.480 0.250 0.0000035 C c.898C be in strong linkage disequilibrium (r =0.70) with the most fi

a signi cantly associated SNP from our previous GWAS, located

a a a a a 2 in the first intron (rs4664308; P=4.3310 28). The WMGL motif (representing tryptophan, methionine, glycine, and leucine) is one of the typical, highly conserved tetrapeptides, found as parts of CTLD b2andb3 strands of the PLA2 receptor (Figure 1). It can be hypothesized that a single change Sequence variants in in these residues could result in changes in binding specificity, 2+ fi 10,11

cantly associated with iMN. Ca sensitivity, or af nity. bronectin type II domain; CRD, cysteine-rich domain; UTR, untranslated region. 160840584 rs149133741160862197 C rs149960520160873180 A/C C rs140427239 0.005 C/T T 0 C/T 0.008 0160889495 rs149256089 A A/G 0.008 0.001 fi fi The other two nonsynonymous SNPs identified in this study b b b b Rare variants. 2 160808075 rs3828323 2 160885442 rs3749117 2 160879259 rs33985939 C C/T 0.094 0.038 0.055 G c.1211G 22 160808076 rs72954858 160833188 rs2715918 G G/A 0.012 0.042 0.014 C c.3315C 2 160901517 rs4665143 2 160885418 rs35771982 2 160919020 rs3749119 2 160918984 rs925409 T T/A 0.074 0.068 0.81 Signi 2 Not reported in 1000 Genomes; allele frequency from Single Nucleotide Polymorphism Database, release 136. 2 2 2 Table 2. Chr Position Name Ref All CEU AF AF Patients Chr, chromosome; position, base pair position based on University of California, Santa Cruz, genome browser versiona human (February 2009) (GRCh37/ b c FNII, genomic reference allele; all, alleles;variation CEU allele AF, frequencies variation in allele controls frequencies and white patients; population ref cDNA, from reference 1000 Genomes allele cDNA release (given 10 that (March 2012); AF patient (rs35771982; p.His300Asp and rs3828323; p.Gly1106Ser) have

680 Journal of the American Society of Nephrology J Am Soc Nephrol 24: 677–683, 2013 www.jasn.org CLINICAL RESEARCH

the carbohydrate recognition domain (CRD), the fibronectin type II domain, and CTLD 1–3.13 Of note, two nonsynonymous common variants identified in this study (p.Met292Val and p.His300Asp) are located in this putative epitope, whereas a third variant, Arg404His, with the same location did not reach statistical significance (Figure 1). Receptors in the mannose receptor family can present with at least two configurations: a straight extended conformation with the N-terminal CRD pointing out- ward from the cell surface and a bent conformation in which the N-terminal CRD is folded back to interact with CTLDs at the middle of the structure. The different configurations affect ligand binding and oligomerization. Anti-PLA2R1 antibodies are directed to a reduction-sensitive epitope,3 which might occur in only one of the two conformations.1,14 The concept of the role of conformational changes in antibody formation was de- veloped by Hudson’s group, who demonstrated that changes in the molecular architecture of the a345-collagen IV autoantigen play a key role in eliciting an autoimmune response in Goodpas- ture syndrome.15 The PLA2R1 sequencing data do not provide a simple genetic basis for iMN as another “conformeropathy.” However, the combined effects of the nonsynonymous variants fi Figure 1. Structure of PLA2R1. The amino acid sequence of signi cantly associated with iMN might alter the conformational PLA2R1 starts with a 20–amino acid signal peptide, followed by changes resulting in subtle alterations of antigenic properties. a large extracellular segment, a transmembrane domain, and Sequence variants may also affect PLA2R1 function, irre- a short intracellular domain containing an endocytosis motif. The spective of its antigenic properties (i.e., binding to antibody). extracellular segment contains a cystine-rich domain at the PLA2R1 is a multifunctional receptor, and its CTLDs are in- N-terminal, followed toward the membrane surface by a fibronectin volved in several functions, such as extracellular matrix orga- type II domain and eight C-type lectin-like domains/carbohydrate nization, endocytosis, complement activation, and cell-to-cell recognition. CD, intracellular domain; COOH, intracellular C termi- interactions.16,17 Alterations in these domains, especially a nal end; CRD, cystine-rich domain (thin vertical lines, S-S bounds); residuechange(p.Met292Val,p.His300Asp)inthefirst fi FNII, bronectin type II domain; CTLD, C-type lectin-like domain; CTLD, may alter these important functions. NH2, extracellular N terminal end; TMD, transmembrane domain. In This study confirmed the association between PLA2R1 (by red, common variants significantly associated with iMN. Rare and new coding variants were found in only nine patients. use of common coding SNPs) and iMN. This may raise the question as to why iMN is rare if these SNPs occur frequently in the general population. We speculate that a combination of previously been shown, in candidate gene association studies known common variants may result in rare haploblocks con- that tested only single SNPs, to be associated with iMN by ferring susceptibility to iMN and that a contingency of cau- groups from Taiwan (p.His300Asp)6 and Korea (p.His300Asp, sality on more than one predisposing gene may occur, given p.Gly1106Ser).7 Both SNPs reached significance for allelic as- our previous study showing that iMN is independently asso- 2 sociation with iMN in the present study (P=1.6310 3 for ciated with variants in at least one molecule encoded in the 2 rs3828323 and P=5.7310 6 for rs35771982) and might be HLA region on the short arm of chromosome 6.5 Also, a rare particularly relevant also because they were identified in eth- combination of (relatively) common events (genetic variants nically distant populations. p.M292V, p.His300Asp, and or environmental factors) might constitute the pathogenetic p.Gly1106Ser are responsible for important amino acid sub- trigger. For example, a variant enhancing the podocyte ex- stitutions that may change the antigenicity or biologic pro- pression of PLA2R1 in the context of an infectious event perties of PLA2R1 as a receptor. might coincide with a variant modifying the conformation Little is known about the process of antibody formation and of the receptor (possibly under specific chemical or environ- the pathogenic epitopes of PLA2R1. PLA2R1 is a type I single- mental constrains, such as heavy-metal presence and drugs). pass transmembrane protein member of the mannose receptor This novel structure would then be recognized as an intrinsic family, which is a member of the C-type lectin superfamily, epitope and presented by an HLA gene (e.g., HLA-DQ recep- itself one of the families of membrane-bound pattern recog- tor) correspondingly shaped by another variant to the effector nition receptors (Figure 1).10–12 It has been suggested that the arm of the immune system, triggering the autoimmune re- potential autoantigenic region resides in the N-terminal part sponse, which would lead to iMN-type lesions. of the receptor because human anti-PLA2R1 antibodies Thus,webelievetherarityofiMNcanbeexplainedbythe recognized a shortened form of the receptor, consisting of underlying genetic basis we are proposing (e.g.,theco-occurrence

J Am Soc Nephrol 24: 677–683, 2013 PLA2R1 Sequence Variants in iMN 681 CLINICAL RESEARCH www.jasn.org of three or more relatively frequent independent events [i.e., available for all patients. Furthermore, presence of anti-PLA2R1 anti- presence of one or more PLA2R1 risk alleles, possibly com- bodies was evaluated. Because this was a cross-sectional cohort, sera bined in a rare haplotype; similarly, occurrence of HLA risk or biopsy specimens were not available for all patients. It is well alleles or haploblocks; and environmental trigger factors]) established that antibodies can disappear after presentation.18 Anti- leading to a rare outcome event (onset of iMN). Further studies PLA2R1 positivity was therefore based on the presence of antibodies (genetic and functional) are needed to elucidate this potential in the serum or the presence of the antigen in deposits in a renal mechanism. biopsy specimen.19 The assays for measuring anti-PLA2R1 antibodies Our study has several limitations. We selected patients from and detecting PLA2R1 in kidney biopsy samples have been described the Dutch and French cohort because the British DNA previously.19 collection has no sufficient antibody status linked to it. It is The study was approved by the ethics committees of the participating possible that a minority of the British patients also would have institutions. All Dutch and French patients gave written informed similar or other rare variants. Given the similarity among the consent before participation in the study. three populations, we consider it unlikely that the proportion of novel mutations would be grossly different. Another Sequencing limitation of this study is that causal noncoding variants could DNA was isolated using salt extraction according to a previously have been missed because we sequenced only exons and described method.20 All exons, including all coding exons and splice essential splice sites. Other potential regions of interest, not sites of PLA2R1, were sequenced using genomic DNA. Primers were sequenced in this study, include putative regulatory regions, designed using Primer3.21 PCR primers and conditions are given in which might affect the splicing pattern (exonic or intronic Supplemental Table S2. PCR products were purified using Multi- splicing enhancers or silencers) and/or the intensity (quantity) screen filter plates (Millipore, Carrigtwohill, Cork, Ireland). Purified or (temporo-spatial) quality of the expression of PLA2R1 iso- products were used for Sanger sequence analysis (unidirectional) forms (i.e., promoter region, transcription factor–binding with a dye-termination chemistry (BigDye Terminator, version 3) sites, enhancer regions, inhibitor regions, insulator elements, on a 3730 DNA analyzer (Applied Biosystems, Inc., Foster City, CA). matrix attachment regions, micro RNA coding, and/or binding sites). A more comprehensive sequencing approach would Analysis be required to investigate this. The sequences of PLA2R1 were investigated for the presence of In summary, in 95 biopsy-proven cases, including 60 PLA2R1 all genetic variants in comparison with the published reference – antibody positive cases of iMN, we found no evidence for our sequence (mRNA NM_007366.4; protein NP_031392.3). Ethnically initial hypothesis that rare variants within the coding region of matched allele frequencies were obtained from the 1000 Genomes PLA2R1 cause the association between iMN and the PLA2R1 project (release 10, March 2012) and from the Single Nucleotide gene. This observation is important because it could prevent Polymorphism Database, release 136. other investigators from engaging in further labor and cost- Variants were considered to be common (previously reported with intensive sequencing of PLA2R1 exons in the quest for poten- allele frequency .1%), rare (previously reported with allele frequency tial rare pathogenic PLA2R1-coding region variants. Second, ,1%), or novel (not previously reported in 1000 Genome project or the by performing a sequence-based association study of the com- Single Nucleotide Polymorphism Database, release 136). fi plete PLA2R1 coding region, we can now con rm on a single Genotype data of 379 white controls were downloaded from the fi base-pair level the ndings of our previous GWAS, which was 1000 Genomes project (release 10, March 2012; details provided in based predominantly on common intronic SNPs. That is, there Supplemental Table S3).22 Using these data and the data of our co- fi are signi cant associations between informative coding var- hort, association analyses, and statistic analyses were performed using iants within PLA2R1 and iMN. Third, this study suggests that PLINK.23 Significance was conservatively set at P=0.01. the strong genetic association between iMN and PLA2R1 is due PLA2R1 was annotated using the custom track functionality of the to intronic regions harboring rare variants or a combination of Genome Browser at the University of California, Santa Cruz;24 its common variants forming a rare haploblock. domain structure is based on UniProt entry Q13018; C-type lectin domain substructure is based on the published literature.25 CONCISE METHODS

Patients ACKNOWLEDGMENTS The European Membranous Nephropathy consortium enrolled Dutch, French, and British cohorts to establish the relation between The research leading to these results has received funding from the the PLA2R1 gene and the development of iMN. For this study, we European Community’s Seventh Framework Programme under analyzed 95 biopsy-proven patients from the Dutch and French co- grant agreement 2012-305608 “European Consortium for High- horts. In all patients, the diagnosis of iMN was established by renal Throughput Research in Rare Kidney Diseases” (EURenOmics)” biopsy, and secondary causes were excluded according to local rou- (to P.W.M., P.E.B., R.K., J.F.M.W., P.R.), by the David and Elaine tine clinical workup.5 Baseline data on age, sex, and ethnicity were Potter Charitable Foundation (to S.H.P., R.K.), by the Dutch

682 Journal of the American Society of Nephrology J Am Soc Nephrol 24: 677–683, 2013 www.jasn.org CLINICAL RESEARCH

Kidney Foundation (J.M.H., grant KJPB11.021), and by the French binding is sensitive to both calcium and lysine. JBiolChem273: 29241– Foundation for Medical Research (H.D., P.R., grant 2012). The 29246, 1998 French Gn-Progress cohort was supported by grants from the 11. Weis WI, Drickamer K: Structural basis of lectin-carbohydrate recognition. Annu Rev Biochem 65: 441–473, 1996 French Ministry of Health (PHRC AOM 00022), the Ministry of 12. Lambeau G, Ancian P, Barhanin J, Lazdunski M: Cloning and expression Research (Decision d’aide 01P0513), and the Biomedicine Agency of a membrane receptor for secretory phospholipases A2. JBiolChem (AO Recherche et Greffes 2005). 269: 1575–1578, 1994 13. Beck LH: The human membranous antibody: mechanism and moni- toring. ASN Renal Week. November 19, 2010. Available at: http://asn. DISCLOSURES sclivelearningcenter.com/index.aspx?PID=1840&SID=80319. Accessed July 23, 2012 None. 14. Debiec H, Nauta J, Coulet F, van der Burg M, Guigonis V, Schurmans T, de Heer E, Soubrier F, Janssen F, Ronco P: Role of truncating mutations in MME gene in fetomaternal alloimmunisation and antenatal glo- REFERENCES merulopathies. Lancet 364: 1252–1259, 2004 15. Pedchenko V, Bondar O, Fogo AB, Vanacore R, Voziyan P, Kitching AR, Wieslander J, Kashtan C, Borza DB, Neilson EG, Wilson CB, Hudson 1. Ronco P, Debiec H: Pathogenesis of membranous nephropathy: Recent BG: Molecular architecture of the Goodpasture autoantigen in anti- – advances and future challenges. Nat Rev Nephrol 8: 203 213, 2012 GBM nephritis. NEnglJMed363: 343–354, 2010 2. Debiec H, Guigonis V, Mougenot B, Decobert F, Haymann JP, 16. Weis WI, Taylor ME, Drickamer K: The C-type lectin superfamily in the Bensman A, Deschênes G, Ronco PM: Antenatal membranous glo- immune system. Immunol Rev 163: 19–34, 1998 merulonephritis due to anti-neutral endopeptidase antibodies. NEngl 17. Day AJ: The C-type carbohydrate recognition domain (CRD) super- – JMed346: 2053 2060, 2002 family. Biochem Soc Trans 22: 83–88, 1994 3. Beck LH Jr, Bonegio RG, Lambeau G, Beck DM, Powell DW, Cummins TD, 18. Hofstra JM, Beck LH Jr, Beck DM, Wetzels JF, Salant DJ: Anti-phospholipase Klein JB, Salant DJ: M-type phospholipase A2 receptor as target antigen A₂ receptor antibodies correlate with clinical status in idiopathic membra- – in idiopathic membranous nephropathy. NEnglJMed361: 11 21, 2009 nous nephropathy. Clin J Am Soc Nephrol 6: 1286–1291, 2011 ₂ 4. Hofstra JM, Wetzels JF: Anti-PLA R antibodies in membranous nephrop- 19. Debiec H, Ronco P: PLA2R autoantibodies and PLA2R glomerular de- – athy: Ready for routine clinical practice? Neth J Med 70: 109 113, 2012 posits in membranous nephropathy. N Engl J Med 364: 689–690, 2011 5. Stanescu HC, Arcos-Burgos M, Medlar A, Bockenhauer D, Kottgen A, 20. Miller SA, Dykes DD, Polesky HF: A simple salting out procedure for Dragomirescu L, Voinescu C, Patel N, Pearce K, Hubank M, Stephens extracting DNA from human nucleated cells. Nucleic Acids Res 16: HA, Laundy V, Padmanabhan S, Zawadzka A, Hofstra JM, Coenen MJ, 1215, 1988 den Heijer M, Kiemeney LA, Bacq-Daian D, Stengel B, Powis SH, 21. Rozen S, Skaletsky H: Primer3 on the WWW for general users and for Brenchley P, Feehally J, Rees AJ, Debiec H, Wetzels JF, Ronco P, biologist programmers. Methods Mol Biol 132: 365–386, 2000 Mathieson PW, Kleta R: Risk HLA-DQA1 and PLA(2)R1 alleles in idio- 22. 1000 Genomes Project Consortium: A map of human genome variation – pathic membranous nephropathy. NEnglJMed364: 616 626, 2011 from population-scale sequencing. Nature 467: 1061–1073, 2010 6. Liu YH, Chen CH, Chen SY, Lin YJ, Liao WL, Tsai CH, Wan L, Tsai FJ: 23. Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MA, Bender D, Association of phospholipase A2 receptor 1 polymorphisms with idio- Maller J, Sklar P, de Bakker PI, Daly MJ, Sham PC: PLINK: A tool set for pathic membranous nephropathy in Chinese patients in Taiwan. JBio- whole-genome association and population-based linkage analyses. med Sci 17: 81, 2010 Am J Hum Genet 81: 559–575, 2007 7. Kim S, Chin HJ, Na KY, Kim S, Oh J, Chung W, Noh JW, Lee YK, Cho JT, 24. Kent WJ, Sugnet CW, Furey TS, Roskin KM, Pringle TH, Zahler AM, Lee EK, Chae DW; Progressive Renal Disease and Medical Informatics Haussler D: The human genome browser at UCSC. Genome Res 12: and Genomics Research (PREMIER) members: Single nucleotide poly- 996–1006, 2002 morphisms in the phospholipase A2 receptor gene are associated with 25. Zelensky AN, Gready JE: The C-type lectin-like domain superfamily. genetic susceptibility to idiopathic membranous nephropathy. Neph- FEBS J 272: 6179–6217, 2005 ron Clin Pract 117: c253–c258, 2011 8. Klein RJ, Zeiss C, Chew EY, Tsai JY, Sackler RS, Haynes C, Henning AK, SanGiovanni JP, Mane SM, Mayne ST, Bracken MB, Ferris FL, Ott J, Barnstable C, Hoh J: Complement factor H polymorphism in age- related macular degeneration. Science 308: 385–389, 2005 See related editorial, “Genetic Variants in Membranous Nephropathy: 9. Segelmark M: Genes that link nephritis to autoantibodies and innate Perhaps a Perfect Storm Rather than a Straightforward Conformeropathy?,” on – immunity. NEnglJMed364: 679 680, 2011 pages 525–528. 10. Graversen JH, Lorentsen RH, Jacobsen C, Moestrup SK, Sigurskjold BW, Thogersen HC, Etzerodt M: The plasminogen binding site of the C-type This article contains supplemental material online at http://jasn.asnjournals. lectin tetranectin is located in the carbohydrate recognition domain, and org/lookup/suppl/doi:10.1681/ASN.2012070730/-/DCSupplemental.

J Am Soc Nephrol 24: 677–683, 2013 PLA2R1 Sequence Variants in iMN 683 SUPPLEMENTARY APPENDIX Supplementary Table S1:

Serum sample biopsy 43 aPLA2R1 + 11 PLA2R1 + 43 positive 2 PLA2R1 - 10 not available 39 aPLA2R1 - 11 PLA2R1 + 11 positive 10 PLA2R1 - 18 not available 13 not available 6 PLA2R1 + 6 positive 2 PLA2R1 - 5 not available 60 positives aPLA2R1 = anti-PLA2R1 antibodies ; PLA2R1 = PLA2R1 in depositions positive = patients are considered to have anti-PLA2R1 related disease if either antibodies were present in serum or PLA2R1 antigen was detectable in renal biopsy (see Methods).

Supplementary Table S2: annealing Exon Forward primer Reverse primer temperature 1 AGTGGGTCTCCAGGGCTC AGGACTGCTGACACACGAAC 64 2 AGATCTGTGGGAAATGGCTG GGTCCAGGCTTCGTGTACTTC 57 3 TGAGTATCTGCCCCAAGAGG CTTCAATATATCAATCCCAGTGC 57 4 GGCCCTGAAGTTTTAGAGATTG TTTTGGGTGAGAGTTTTGGG 57 5 TGAGCCTGCTCTGTTTCTCA TGGCTTACACCCAAATCCTC 57 6 CATGCCTGGACTGTAACTGC CACAGTCCCTTTGACATAACAGA 64 7 CACTAGCTCATCCGGAAAATG GGGGTTCCCTTTATTATGTGG 57 8 TCAAACCATAGCAGGGAAGG CACAGGTGGCCTTCAGTACA 64 9 GCACATGGCAGCTATCAAGA GACAGGGCATCACACATCAC 57 10 AGAAGTGGGTGTAGTTAATGGC GGGTTTCATTTCCCATATAAGC 57 11 TACTCTGGGGCCATTCTGTG CCCTGTTAACCTGTGTTGATCC 57 12 TTTGCCATCTGAGCACAAAG TCTGTCAGGCACTGTTCTGG 57 13 TTCCTCCCACTAGATCAAGCTC GGGAATTTTAAACAACTACTTGGG 57 14 GCATCAGCTGTTTCCTGC GTGCAGTTTTGTCATGGGAG 64 15 AGGGATGCCCTTTTTGATCT CTGTCCCAGGTGTTTCCACT 57 16 GCCTCTCTGGAGCTTCCTCT AAAAGGTCGGGGATAGGTACA 64 17 TGTACCTATCCCCGACCTTTT AAACAATCTCGACGTGCTCA 64 18 GGTTTTCTCAAGATGCTAGGC CACTCATGTCTTTGGTTAGAGAAG 57 19 TGGAGACTGGACATGCAGAA CCCATTTCTGGTGCTGTCTT 64 20 CCTTGCCCAGTTAGCTGTTG TAGAAACGAGGGGCAGTGTC 64 21 CTGTAACACCTTTGGAGGAGG GGTCAAACTGCAAACTCGTG 64 22 AAATAGGATGCAGTGATTACCC GGCACAGAGAGGAATGTGTTT 64 23 TGCGAATAAAGGAATAATGGTG CATGCCACATTTAGCCACAG 57 24 CAACATGGATCATAGTGACAAGG AGGCTGTTCATAAGGTTGGTC 64 25 TCGATGTGGGAATTGTAGCA ATCACCAACCAGACCAGAGG 57 26 CATGAAACCCAATCTTTCAGAG GGTTTCCAAATTCTACACTGGAC 64 27 GCCTTAACTTCTGACGTGCAA AGCCAAGCATCTTCCTGCTA 64 28 AGGGGCTGCTTTTTATGGTT CCAGAATGAAGGTTGCTGGT 64 29 and 30 GTCATGGGCCATAGTTTGCT TGGCATCTGTGCTGTAAAGG 64 All primer sequences given are from 5' to 3'. Underlined primers were used for sequencing. For exon 9 the following primer was used for sequencing (TTCAAATTGATGGTGGAATTGA). PCR was performed using the Amplitaq Gold 360 PCR mix (Applied Biosystems); for the amplification of exon 4, 5 and 7 Amplitaq Gold (Applied Biosystems) was used.

Supplementary Table S3: population of European Ancestry , 1000 Genomes Project, release 10 March 2012

Population Number Utah residents (CEPH) with Northern & Western European ancestry (CEU) 85 Toscani in Italy (TSI) 98 British from England and Scotland (GBR) 89 Finnish from Finland (FIN) 93 Iberian population in Spain (IBS) 14 Total European ancestry (CEU) 379